LDL receptor-GFP fusion proteins: new tools for the characterisation of disease-causing mutations in the LDL receptor gene.

The function of a series of LDL receptor GFP fusion proteins with different, flexible, unstructured spacer regions was analysed. An optimised version of the fusion protein was used to analyse the effect of an LDL receptor mutation (W556S) found in FH patients and characterised as transport defective. In cultured liver cells this mutation was found to inhibit the transport of LDL receptor GFP fusion protein to the cell surface, thus leading to impaired internalisation of fluorescent labelled LDL. Co-localisation studies confirmed the retention of the mutant protein in the endoplasmic reticulum. Wild type (WT) and W556S LDL receptor GFP fusion proteins were expressed in mouse liver by means of hydrodynamic delivery of naked DNA. Two days after injection liver samples were analysed for GFP fluorescence. The WT LDL receptor GFP protein was located on the cell surface whereas the W556S LDL receptor GFP protein was retained in intracellular compartments. Thus, the GFP-tagged LDL receptor protein allows both detailed time lapse analysis and evaluations in animals for the physiological modelling of mutations. This method should be generally applicable in functional testing of gene products for aberrant processing.

Ribosome synthesis in Tetrahymena: a quantitative analysis.

We have performed a detailed quantitative analysis of the transcription and accumulation of ribosomal RNA and ribosomal protein mRNA in the ciliated protozoan Tetrahymena thermophila during changes in growth conditions, and found that: (1) nutritional downshifts lead to a rapid decrease in transcriptional activity whereas nutritional upshifts lead to rapid restoration of transcriptional activity, (2) starvation leads to decreased translation of ribosomal protein mRNA and (3) the rate of ribosomal protein mRNA degradation decreases after a nutritional upshift. We present evidence that the proximal promoters of two ribosomal protein genes and the ribosomal RNA gene compete for binding of nuclear factor(s) in vitro, suggesting that the coordinated regulation of these genes may involve a common set of transcriptional regulators.

Clinical and molecular evidence of abnormal processing and trafficking of the vasopressin preprohormone in a large kindred with familial neurohypophyseal diabetes insipidus due to a signal peptide mutation.

The autosomal dominant form of familial neurohypophyseal diabetes insipidus (adFNDI) is a rare disease characterized by postnatal onset of polyuria and a deficient neurosecretion of the antidiuretic hormone, arginine vasopressin (AVP). Since 1991, adFNDI has been linked to 31 different mutations of the gene that codes for the vasopressin-neurophysin II (AVP-NPII) precursor. The aims of the present study were to relate the clinical phenotype to the specific genotype and to the molecular genetic effects of the most frequently reported adFNDI mutation located at the cleavage site of the signal peptide of AVP-NPII [Ala(-1)Thr]. Genetic analysis and clinical studies of AVP secretion, urinary AVP, and urine output were performed in 16 affected and 16 unaffected family members and 11 spouses of a Danish adFNDI kindred carrying the Ala(-1)Thr mutation. Mutant complementary DNA carrying the same mutation was expressed in a neurogenic cell line (Neuro2A), and the cellular effects were studied by Western blotting, immunocytochemistry, and AVP measurements. The clinical studies showed a severe progressive deficiency of plasma and urinary AVP that manifested during childhood. The expression studies demonstrated that the Ala(- 1)Thr mutant cells produced 8-fold less AVP than wild-type cells and accumulated excessive amounts of 23-kDa NPII protein corresponding to uncleaved prepro-AVP-NPII. Furthermore, a substantial portion of the intracellular AVP-NPII precursor appeared to be colocalized with an endoplasmic reticulum antigen (Grp78). These results provide independent confirmation that this Ala(-1)Thr mutation produces adFNDI by directing the production of a mutant preprohormone that accumulates in the endoplasmic reticulum, because it cannot be cleaved from the signal peptide and transported to neurosecretory vesicles for further processing and secretion.

Cloning and characterization of the gene encoding the highly expressed ribosomal protein l3 of the ciliated protozoan Tetrahymena thermophila. Evidence for differential codon usage in highly expressed genes.

We have cloned and characterized the cDNA and the macronuclear genomic copy of the highly conserved ribosomal protein (r-protein) L3 of Tetrahymena thermophila. The r-protein L3 is encoded by a single copy gene interrupted by one intron. The organization of the promoter region exhibits features characteristic of ribosomal protein genes in Tetrahymena. The codon usage of the L3 gene is highly biased. A thorough analysis of codon usage in Tetrahymena genes revealed that genes could be categorized into two classes according to codon usage bias. Class A comprises r-protein genes and a number of other highly expressed genes. Class B comprises weakly expressed genes such as the conjugation induced CnjB and CnjC genes, but surprisingly, this class also contains abundantly expressed genes such as the genes encoding the surface antigens SerH3 and SerH1. Codon usage is slightly more restricted in class A than in class B, but both classes exhibit distinct and different codon usage biases. Class A genes preferentially use C and U in the silent third codon positions, whereas class B genes preferentially use A and U in the silent third codon positions. The analysis suggests that two different strategies have been employed for optimization of codon usage in the A+T-rich genome of Tetrahymena.

Normolipidemia and hypercholesterolemia in persons heterozygous for the same 1592 + 5G --> A splice site mutation in the low-density lipoprotein receptor gene.

In the present study, we have characterized a unique splice donor G to A substitution in the moderately conserved + 5 position in intron 10 of the low-density lipoprotein (LDL) receptor gene. In two Danish families, carriers of the 1592 + 5G --> A mutation display a clinical phenotype ranging from healthy normocholesterolemic persons to classical heterozygous familial hypercholesterolemia (FH) patients. Reverse transcription-polymerase chain reaction (RT-PCR) of RNA from Epstein Barr virus (EBV)-transformed lymphoblasts obtained from members of both families demonstrated abnormal splicing generating two aberrant mRNAs due to either alternative splicing and skipping of exon 10 or activation of a cryptic splice site in intron 10 inserting 66 intronic base pairs. These abnormally spliced mRNAs were predicted to encode two abnormal receptor proteins containing an in-frame deletion of 75 amino acids and an insertion of 22 novel amino acids, respectively. Results obtained by immunofluorescence staining, flow cytometry, and confocal microscopy of transfected Chang and COS-7 cells expressing normal and mutant LDL receptors were compatible with nearly complete retention of the mutant proteins in the endoplasmic reticulum. Quantitative measurements of LDL receptor mRNAs from EBV-transformed lymphoblasts, however, did not reveal any significant differences in variant mRNA contents between mutation carriers in the families that could be related to degree of hypercholesterolemia.

Structural organization of the human short-chain acyl-CoA dehydrogenase gene.

Short-chain acyl-CoA dehydrogenase (SCAD) is a homotetrameric mitochondrial flavoenzyme that catalyzes the initial reaction in short-chain fatty acid beta-oxidation. Defects in the SCAD enzyme are associated with failure to thrive, often with neuromuscular dysfunction and elevated urinary excretion of ethylmalonic acid (EMA). To define the genetic basis of SCAD deficiency and ethylmalonic aciduria in patients, we have determined the sequence of the complete coding portion of the human SCAD gene (ACADS) and all of the intron-exon boundaries. The SCAD gene is approximately 13 kb in length and consists of 10 exons. Four polymorphic sites have previously been detected by sequencing of cDNA from fibroblasts of patients excreting elevated amounts of EMA. Three of these polymorphisms (321T/C, 990C/T, 1260G/C) are silent variants, while a 625G/A polymorphism results in an amino acid replacement and has been shown to be associated with ethylmalonic aciduria. From analysis of 18 unrelated Danish families, we show that the four SCAD gene polymorphisms constitute five allelic variants of the SCAD gene, and that the 625A variant together with the less frequent variant form of the three other polymorphisms (321C, 990T, 1260C) constitutes an allelic variant with a frequency of 22% in the general Danish population. Using fluorescence in-situ hybridization, we confirm the localization of the human SCAD gene to the distal part of Chromosome (Chr) 12 and suggest that the SCAD gene is a single-copy gene. The evolutionary relationship between SCAD and five other members of the acyl-CoA dehydrogenase family was investigated by two independent approaches that gave similar phylogenetic trees.

A common W556S mutation in the LDL receptor gene of Danish patients with familial hypercholesterolemia encodes a transport-defective protein.

In a group of unrelated Danish patients with familial hypercholesterolemia (FH) we recently reported two common low-density lipoprotein (LDL) receptor mutations, W23X and W66G, accounting for 30% of the cases. In this study, we describe another common LDL receptor mutation, a G to C transition at cDNA position 1730 in exon 12, causing a tryptophan to serine substitution in amino acid position 556 (W556S). In the Danish patients, the W556S mutation was present in 12% of 65 possible mutant alleles. The pathogenicity of the W556S mutation, which is located in one of the five conserved motifs Tyr-Trp-Thr-Asp in the epidermal growth factor homology region, was studied in transfected COS-7 cells expressing normal and mutant LDL receptor cDNAs. Results obtained by immunofluorescence flow cytometry and confocal microscopy, as well as by immunoprecipitation, were compatible with complete retention of the mutant protein in the endoplasmic reticulum. The transport-defective W556S mutation and the W23X and W66G mutations seem to account for about 40% of the LDL receptor defects in Danish families with FH.

Two mutations in the same low-density lipoprotein receptor allele act in synergy to reduce receptor function in heterozygous familial hypercholesterolemia.

Mutations in genes are not necessarily pathogenic. Expression of mutant genes in cells can therefore be required to demonstrate that mutations in fact disturb protein function. This applies especially to missense mutations, which cause an amino acid to be replaced by another amino acid. In the present study of two families with familial hypercholesterolemia in the heterozygous form, we found two mutations in the same allele of the low-density lipoprotein (LDL) receptor gene: a missense Asn543. His mutation (N543H) in exon 11, and an in-frame 9-bp deletion (2393del9) in exon 17. The two mutations were identified in heterozygous FH index patients in whom no other pathogenic mutations were detected by SSCP analysis of the remaining 16 exons and the promoter region. Both mutations cosegregated with hypercholesterolemia within the families. Each of these mutations had little or no effect on receptor function in transfected COS cells, but when both mutations were present simultaneously, receptor function, as assessed by flow cytometric measurement of fluorescent LDL uptake in cells, was reduced by 75%. Immunostainable receptors on the cell surface were decreased by 80% as measured by flow cytometry. The two mutations therefore acted in synergy to affect receptor function, possibly during intracellular receptor transport, since Northern blot analysis suggested that mRNA levels were unaffected. Without screening of the entire coding regions of the gene, the synergistic action of these two LDL receptor mutations would not have been detected.

Changes in moral reasoning and the teaching of medical ethics.

Courses in medical ethics are becoming an integral part of many medical school curricula in Europe. At the medical school of the University of Copenhagen, a course on philosophy of medicine has been compulsory for all medical students since 1988. The effect of such courses on the ethical awareness and reasoning of medical students is not well understood and we have therefore found it of interest to study the effects of the Copenhagen course. For the study, we used a Danish version of the Defining Issues Test (DIT) which measures development in moral reasoning (Rest J R, 1979 Development in Judging Moral Issues. University of Minnesota Press, Minneapolis). The study was conducted as a pre- and post-test study without a control group, and the subjects were all medical students attending the course in the autumn of 1993. The results show that moral reasoning scores measured by the DIT increase significantly, and we argue that this increase can only be explained as an effect of the course.

The function of acyl-CoA-binding protein (ACBP)/diazepam binding inhibitor (DBI).

Acyl-CoA-binding protein has been isolated independently by five different groups based on its ability to (1) displace diazepam from the GABAA receptor, (2) affect cell growth, (3) induce medium-chain acyl-CoA-ester synthesis, (4) stimulate steroid hormone synthesis, and (5) affect glucose-induced insulin secretion. In this survey evidence is presented to show that ACBP is able to act as an intracellular acyl-CoA transporter and acyl-CoA pool former. The rat ACBP genomic gene consists of 4 exons and is actively expressed in all tissues tested with highest concentration being found in liver. ACBP consists of 86 amino acid residues and contains 4 alpha-helices which are folded into a boomerang type of structure with alpha-helices 1, 2 and 4 in the one arm and alpha-helix 3 and an open loop in the other arm of the boomerang. ACBP is able to stimulate mitochondrial acyl-CoA synthetase by removing acyl-CoA esters from the enzyme. ACBP is also able to desorb acyl-CoA esters from immobilized membranes and transport and deliver these for mitochondrial beta-oxidation. ACBP efficiently protects acetyl-CoA carboxylase and the mitochondrial ADP/ATP translocase against acyl-CoA inhibition. Finally, ACBP is shown to be able to act as an intracellular acyl-CoA pool former by overexpression in yeast. The possible role of ACBP in lipid metabolism is discussed.

Genome organization and expression of the rat ACBP gene family.

Acyl-CoA-Binding Protein (ACBP)/Diazepam-Binding Inhibitor (DBI) is a 10 kD protein which has been implicated in a surprisingly large number of biochemical functions. We have unambiguously demonstrated that ACBP binds acyl-CoA esters with high affinity and in vivo functions as an acyl-CoA ester pool former. We have molecularly cloned and characterized the rat ACBP gene family which comprises one expressed and four processed pseudogenes. One of these was shown to exist in two allelic forms. A comprehensive computer-aided analysis of the promoter region of the expressed ACBP gene revealed that it exhibits all the hallmarks of typical housekeeping genes. In addition, the promoter region harbors a number of potential tissue specific cis-acting elements that may in part regulate the level of ACBP expression in specialized cells.

Acyl-CoA-binding protein/diazepam-binding inhibitor gene and pseudogenes. A typical housekeeping gene family.

Acyl-CoA-binding protein (ACBP) is a 10 kDa protein isolated from bovine liver by virtue of its ability to bind and induce the synthesis of medium-chain acyl-CoA esters. Surprisingly, it turned out to be identical to a protein named diazepam-binding Inhibitor (DBI) claimed to be an endogenous modulator of the GABAA receptor in brain membranes. ACBP/DBI, or proteolytically derived polypeptides of ACBP/DBI, have also been implicated in the control of steroidogenesis in mitochondria and glucose-stimulated insulin secretion. Thus, it appears that ACBP/DBI is a remarkable, versatile protein. Now we have molecularly cloned and characterized the ACBP/DBI gene family in rat. The rat ACBP/DBI gene family comprises one expressed gene and four processed pseudogenes of which one was shown to exist in two allelic forms. The expressed gene is organized into four exons and three introns. There is a remarkable correspondence between the structural modules of ACBP/DBI as determined by 1H nuclear magnetic resonance spectroscopy and the exon-intron architecture of the ACBP/DBI gene. Detailed analyses of transcription of the ACBP/DBI gene in brain and liver were performed to map transcription initiation sites and to examine if transcripts from the ACBP/DBI gene were subject to alternative processing. In both brain and liver, transcription is initiated from two major and multiple minor initiation sites. No evidence for alternative splicing was obtained. The promoter region of the ACBP/DBI gene is located in a CpG island and lacks a canonical TATA box. Thus, the ACDB/DBI gene exhibits all the hallmarks of a typical housekeeping gene.

Tetrahymena gene encodes a protein that is homologous with the liver-specific F-antigen and associated with membranes of the Golgi apparatus and transport vesicles.

The F-antigen is a prominent liver protein which has been extensively used in studies on natural and induced immunological tolerance. However, its intracellular localization and biological function have remained elusive. It has generally been assumed that the F-antigen is confined phylogenetically to vertebrates. Now we have cloned and characterized a gene from the ciliated protozoan Tetrahymena thermophila encoding a protein which clearly is homologous with the rat F-antigen. The coding region of the Tetrahymena F-antigen (TF-ag) gene specifies a 46,051 M(r) protein and is interrupted by three introns. In accordance with the predicted molecular mass of the TF-ag protein, antibodies raised against a cro-lacZ'-TF-ag fusion protein specifically recognized a 45,000 M(r) protein in Western blots of total T. thermophila protein. Immunoelectron microscopy demonstrated that the TF-ag is associated with membranes of the Golgi apparatus and transport vesicles pointing to a role of TF-ag in membrane trafficking. Transcription of the TF-ag gene, as determined by run-on analyses, was only detectable in growing cells, and following transfer to starvation condition pre-existing TF-ag mRNA was rapidly degraded. The abundance of the TF-ag protein, however, declined only moderately during prolonged periods of starvation demonstrating that extensive release of the TF-ag did not take place. In combination these results suggest that the TF-ag protein is a recycled constituent of the intracellular membrane network in T. thermophila.

Tetrahymena thermophila acidic ribosomal protein L37 contains an archaebacterial type of C-terminus.

We have cloned and characterized a Tetrahymena thermophila macronuclear gene (L37) encoding the acidic ribosomal protein (A-protein) L37. The gene contains a single intron located in the 3'-part of the coding region. Two major and three minor transcription start points (tsp) were mapped 39 to 63 nucleotides upstream from the translational start codon. The uppermost tsp mapped to the first T in a putative T. thermophila RNA polymerase II initiator element, TATAA. The coding region of L37 predicts a protein of 109 amino acid (aa) residues. A substantial part of the deduced aa sequence was verified by protein sequencing. The T. thermophila L37 clearly belongs to the P1-type family of eukaryotic A-proteins, but the C-terminal region has the hallmarks of archaebacterial A-proteins.

Induction of acyl-CoA-binding protein and its mRNA in 3T3-L1 cells by insulin during preadipocyte-to-adipocyte differentiation.

The induction of acyl-CoA-binding protein (ACBP) and ACBP mRNA was investigated in 3T3-L1 cells during growth and insulin-induced differentiation. The level of ACBP relative to both total soluble protein and DNA increased during insulin-stimulated conversion of 3T3-L1 cells from preadipocytes into fully developed adipocytes. So did the total rate of lipogenesis, as measured by incorporation of [1-14C]acetate. A similar increase in ACBP mRNA relative to total RNA was observed. These results therefore suggest that ACBP plays a specific role in the lipogenic process. However, this role might be indirect, as the increase in lipogenesis preceded the increase in ACBP. The significance of this finding is discussed.

Structure and evolution of the Tetrahymena thermophila gene encoding ribosomal protein L21.

The single-copy gene encoding ribosomal protein (r-protein) L21 in the macronucleus of the ciliate Tetrahymena thermophila was cloned and characterized. Sequencing of the L21 gene and a corresponding cDNA clone showed the gene to contain three introns, all located in the 3' half of the coding region. Primer extension and nuclease protection analyses revealed five transcription start points (tsp) 56-73 nucleotides upstream from the start codon. The uppermost tsp mapped to the first T in a sequence, TATAA, often found at tsp in T. thermophila. A comparison of amino acid sequences of r-proteins revealed that T. thermophila L21 belongs to a family of r-proteins that have been conserved in eubacteria, archaebacteria and eukaryotes. On the basis of structural and functional considerations, we suggest that the eukaryotic, like the prokaryotic, members of this protein family interact with the 5S RNA complex in ribosomes.

The hok killer gene family in gram-negative bacteria.

The seven members of the hok killer gene family in Gram-negative bacteria are described here. The members of this gene family have been sequenced and include hok/sok from plasmid R1, flm and srnB from plasmid F, pnd from plasmids R483 and R16, and gef and relF, which are located on the Escherichia coli chromosome. The killer proteins encoded by these loci are highly toxic polypeptides of 50 to 52 amino acids. The proteins kill the cells from the inside by interfering with a vital function in the cell membrane. On the basis of their relatedness, the killer proteins and their corresponding loci are divided into four subfamilies. The members of one subfamily, hok/sok and flm, mediate plasmid maintenance by killing plasmid-free cells. The pnd and srnB subfamilies were discovered through their abilities to cause membrane damage and degradation of stable RNA. gef and relF, which constitute the chromosomal subfamily, were found because of their sequence similarity at the DNA and protein levels with other members of the hok gene family. However, no function has been described for the proteins belonging to this subfamily. Although the four subfamilies are distantly related in terms of DNA and protein sequence similarity, the overall genetic organization of the different loci has been well conserved during evolution. The expression of all of the members of the hok gene family is regulated post-transcriptionally. Thus, the expression of the hok and flm genes is regulated by small antisense RNAs that inhibit the translation of the stable hok and flm mRNAs. On the basis of structural and functional similarities, we suggest that each of the related plasmid-encoded killer genes is regulated by antisense RNAs. The conservation of this widespread gene family in Gram-negative bacteria suggests that the genes are important to the genomes that carry them.

Unusual ciliate-specific codons in Tetrahymena mRNAs are translated correctly in a rabbit reticulocyte lysate supplemented with a subcellular fraction from Tetrahymena.

The codon usage of Tetrahymena thermophila and other ciliates deviates from the 'universal genetic code' in that UAA and probably UAG are not translational termination signals but code for glutamine. Therefore, translation in vitro of mRNA from Tetrahymena in a reticulocyte lysate is prematurely terminated if a UAA or UAG triplet is present in the reading frame of the mRNA. We show that the addition of a subcellular fraction from Tetrahymena thermophila enables a rabbit reticulocyte lysate to translate Tetrahymena mRNAs into full-sized proteins. The activity of the subcellular fraction is shown to depend on the combined function of a protein component(s) and a tRNA(s). The subcellular fraction is easily prepared and its usefulness for the identification of isolated mRNAs from Tetrahymena by their translation products in vitro is demonstrated.

An intron in a ribosomal protein gene from Tetrahymena.

We have cloned and sequenced a single copy gene encoding a ribosomal protein from the ciliate Tetrahymena thermophila. The gene product was identified as ribosomal protein S25 by comparison of the migration in two-dimensional polyacrylamide gels of the protein synthesized by translation in vitro of hybrid-selected mRNA and authentic ribosomal proteins. The proteins show strong homology to ribosomal protein S12 from Escherichia coli. The coding region of the gene is interrupted by a 979-bp intron 68 bp downstream of the translation start. This is the first intron in a protein encoding gene of a ciliate to be described at the nucleotide sequence level. The intron obeys the GT/AG rule for splice junctions of nuclear mRNA introns from higher eukaryotes but lacks the pyrimidine stretch usually found in the immediate vicinity of the 3' splice junction. The structure of the intron and the fact that it is found together with the well described self-splicing rRNA intron is discussed in relation to the evolution of RNA splicing.

The increases in the rates of synthesis of ribosomal proteins and ribosomal RNA during refeeding of starved Tetrahymena cells are not dependent on DNA replication.

We have analysed the effects of an inhibition of DNA replication by hydroxyurea on the synthesis of ribosomal proteins (r-proteins) and ribosomal RNA (rRNA) in Tetrahymena cells resuming growth after long-term starvation. The coordinate regulation of the synthesis of individual r-proteins and their increased rate of synthesis during refeeding are not impaired by inhibition of DNA replication. Moreover, the presence of hydroxyurea does not prevent an increase in the rate of synthesis of rRNA around 70-80 min after refeeding. Previously, this increase was claimed to be gene dose-dependent. Up to 180 min after refeeding, the synthesis of r-proteins appears to be closely coupled with that of rRNA and proceed in stoichiometric balance, irrespective of whether hydroxyurea is present or not. After 180 min of refeeding in the presence of hydroxyurea, this stoichiometric balance breaks down, and the rate of synthesis of r-proteins clearly exceeds that of the rRNA synthesis.